This isn't quite dumping. It was more about achieving the scale necessary to make it usable. If there are no drivers, there are no customers. With no customers, who would drive?
It is indeed dumping. Scale out so big that previous competitors can't keep up, then squeeze hard shaking out other competitors(taxis/rentals) as well as your own minions(a.k.a. gig economy workers) for blood.
> If there are no drivers, there are no customers. With no customers, who would drive?
No worries, Waymo is coming soon to kill Uber as well as everyone Uber took out of business.
As you mentioned, the data processing inequality[1] applies here, but I imagine synthetic data could help training squeeze out more from the existing data.
It's neat how a longer "digestive tract" loses entropy, but can make up for it by making more sense of things. It's akin to adding a NN layer, to a more computationally-intensive lossy compression algorithm, or to asking a LLM to explain the problem domain and the relevant variables (populating attention) before getting to the point.
It's probably true for people too. Instead of asking an expert for an opinion right away, ask them to discuss the options out loud first.
There are probably a lot of applications where the LLM could rely more on data that's supplied to it just-in-time in the context window, and less on specialist knowledge from its training set.
Also, "natural" data taken from the Internet is probably quite inefficient as training material. It's going to have a lot of duplication. You only need each fact once to be able to synthesize more examples of it.
As long as it cannot be oppressive to the majority of the population, and until you are free to emigrate taking your belongings with you, calling "oppressive" a modern democratic country is a huge lack of respect to those people who lived their entire life with no choice at all.
> I've seen people say that apartments that are being provided to the homeless for free should have granite countertops or GTFO
This is a common problem in American policy. You're going to have the thing and it's going to be nice, or you're going to be priced out of having it at all.
See minimum floor space regulations, (non-fire related) residential occupancy limits, regulations requiring certain coverage in medical insurance plans, etc. There is a reason that short term insurance plans (STLDIs) that are not subject to as many regulations are much cheaper.
I didn't mean to imply there was anything like voluntary consent involved. I'm on your side here. But there is popular support for this program which wouldn't exist if it were presented purely as a wealth transfer with no upside for those forced to pay in. They gloss over the fact that paying SS taxes doesn't formally entitle you to any future benefits, but in practice cancelling it without offering some compensation to those who paid so much in would amount to political suicide.
This brings up and interesting question: If more people enter the work force and pay taxes every year, how is SS (as im told) "drying up"? I don't doubt that it's dwindling, however, what happened?
In 1945 there were 42 workers paying SS taxes for every retiree collecting benefits. For equal pay / retirement income after SS taxes, each worker only needs to provide 2.3%.
Today (actually since 2009) that ratio is three workers per retiree. Each worker has to provide 33%. By 2050 the ratio is projected to fall to two-to-one.
Mostly people are living longer. When SS first started only 55% of males and 60% of females who made it to age 21 would have survived their working-age years to retire at age 65 in 1940. By 1990 that figure had risen to 72% and 83%, respectively. Life expectancy after retirement also increased, from 12.7 (M) or 14.7 (F) years in 1940 to 15.3 or 19.6 years in 1990. (https://www.ssa.gov/history/lifeexpect.html)
There's a culture of forced retirement there too, so it's not at all surprising. Pay generally increases with age so Japanese companies want people to retire.
> My fallacy is that there is a billion things that I am good at, but I don't consider them difficult (how could they be, if I am good at them)
I see you've rediscovered Rothbard's law: "People tend to specialise in what they're worst at." (Because anything too easy seems like it's not worth doing)
reply